From mboxrd@z Thu Jan 1 00:00:00 1970 From: Hans Reiser Subject: Re: batched write Date: Mon, 19 Jun 2006 13:39:19 -0700 Message-ID: <44970B77.6030906@namesys.com> References: <44736D3E.8090808@namesys.com> <20060608121006.GA8474@infradead.org> <1150322912.6322.129.camel@tribesman.namesys.com> <20060617100458.0be18073.akpm@osdl.org> <4494411B.4010706@namesys.com> <87ac8an21r.fsf@hades.wkstn.nix> <449668D1.1050200@namesys.com> <4496D34F.4010007@namesys.com> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Cc: Miklos Szeredi , nix@esperi.org.uk, akpm@osdl.org, vs@namesys.com, hch@infradead.org, Reiserfs-Dev@namesys.com, Linux-Kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org, drepper@redhat.com Return-path: Received: from rwcrmhc11.comcast.net ([216.148.227.151]:1975 "EHLO rwcrmhc11.comcast.net") by vger.kernel.org with ESMTP id S1751151AbWFSUjT (ORCPT ); Mon, 19 Jun 2006 16:39:19 -0400 To: Akshat Aranya , vs In-Reply-To: Sender: linux-fsdevel-owner@vger.kernel.org List-Id: linux-fsdevel.vger.kernel.org Akshat Aranya wrote: > On 6/19/06, Miklos Szeredi wrote: > >> > I would think that batched write is pretty essential then to FUSE >> > performance. >> >> Well, yes essential if the this is the bottleneck in write throughput, >> which is most often not the case, but sometimes it is. >> > > I can vouch for this. I did some experiments with an example FUSE > filesystem that discards the data in userspace. Exporting such a > filesystem over NFS gives us 80 MB/s writes when FUSE is modified to > write with 32K block sizes. With the standard FUSE (4K writes), we > get closer to 50 MB/s. The ratios of 4k performance / large write performance are amusingly similar for reiser4 and FUSE even though the filesystems and absolute performance are totally different. The principle is the same it seems for both filesystems. Vladimir, the benchmarks, please send them..... Hans